I just attended the 2018 Michael, Marcia, and Christa Parseghian Scientific Conference for Niemann-Pick Type C (NPC) research. My mind is still spinning from all the productive interactions and conversations I had over three session-packed days at The Westin La Paloma Resort in the Catalina Foothills north of Tucson. In addition to taking in all the science, it was fun to catch up with regulars like Rich Taylor and Kevin Vaughan from Notre Dame, Suzanne Pfeffer from Stanford, Thomas Kirkegaard from Orphazyme, as well as well engage NPC newcomers like Roberto Zoncu from Berkeley, just to name a few of the 140+ attendees – a conference record. It was also exciting to meet erstwhile attendees returning after a hiatus, such as Yiannis Ioannou from Mount Sinai. It goes without saying but the conference would not be the same without the active participation of NPC families like Chris and Pam Andrews from Texas, Cheryl and Paul Marcogliese from Canada, and, of course, Cindy and Mike Parseghian.
In this post I share my reflections on this year’s meeting, on the arc of NPC research over the five years that I’ve attended this meeting, and on being a member of the NPC rare disease community.
Three big takeaways: clinical trials, biomarkers, 2nd generation therapies
To my mind there were three highlights of the 2018 meeting. First, there was palpable anticipation for the final results of two ongoing Ph2/3 clinical trials for experimental NPC therapies: hydroxypropyl-beta-cyclodextrin (cyclodextrin) and arimoclomol. Both of these trials began in earnest around the time I attended my first Parseghian meeting in 2014, where almost every talk was about cyclodextrin. The rare disease NewCo accelerator Cydan launched Vtesse in 2014 to run the cyclodextrin trial. (Last year Vtesse was acquired by Sucampo, which itself earlier this year was acquired by Mallinckrodt). Based on the Ph1/2 cyclodextrin trial data published last year in The Lancet, as well as the Ph2/3 cyclodextrin trial data showing stabilization of NPC Clinical Severity Score progression presented at this year’s meeting, I would be shocked if cyclodextrin isn’t approved by FDA later this year or early next year. Cyclodextrin works in every NPC disease model tested from nematodes and flies as we showed at Perlara, to mice and cats. The crowning strength of cyclodextrin is its mechanism of action that bypasses NPC1 and NPC2 and drives cholesterol out of the endocytic pathway by mass action effects. The liabilities of cyclodextrin have been evident from the start: intrathecal route of administration due to lack of brain penetration and oral bioavailability, and ototoxicity (high-frequency hearing loss). If all the animal data showing that dosing cyclodextrin earlier in life is better, do parents risk language acquisition and development in NPC toddlers?
Arimoclomol, an orally bioavailable small molecule activator (“amplifier”) of the protein chaperone HSP70 that gets into the brain, doesn’t have those liabilities. The therapeutic thesis for arimoclomol is that direct activation of chaperones like HSP70 rescues mis-folded NPC1 proteins and leads to the cytoprotective recruitment of chaperones to lysosomal membranes. The arimoclomol trial is sponsored by the Danish biotech company Orphazyme, which has been tight-lipped about their trial. However, they did present data on a new blood-based pharmacodynamic marker for NPC. In fact, two other groups presented data on bile acid derivatives as diagnostic biomarkers. Along those lines, the Texas-based NPC patient advocacy group Firefly Fund presented their ambitious newborn screening plan to get NPC1 added to the RUSP — Recommended Uniform Screening Panel.
The cyclodextrin and arimoclomol trials have each enrolled around 50 patients. On the one hand, everyone knows there are only so many NPC patients known to medicine to enroll in clinical trials. On the other hand, everyone also knows there are lots of undiagnosed NPC kids around the world. Second-generation NPC therapies are still in preclinical and discovery stages, but the community is already fretting about the availability of patients for future trials and the need for curative therapies that reverse disease versus slow progression. And no one was really talking aloud about the other elephant in the room: how much will the first NPC drugs cost? An NPC parent did ask the CSO from Mallinckrodt what VTS-270 (their name for cyclodextrin) will cost. After an initial dodge, the CSO conceded that it will price in the range of comparable orphan products, presumably between $200,000 and $400,000 per patient per year.
What does NPC1 actually do?
The NPC1 gene was cloned in 1997 but in 2018 we still don’t fully understand how the NPC1 protein works. Is it a transporter? Is it a pore? Is it a cholesterol sensor? Is it all of those things wrapped up into one protein? Is it only involved in cholesterol trafficking? The most important question not only for the NPC community but also for any rare community is how long should it take to understand a disease gene and prove a theory of the disease? In the case of NPC1, has 21 years been too long or just right given the complexity of this gene’s biology? I don’t know the answer to this question but I do know that there was a sense at this year’s meeting that multiple independent strands of NPC1 basic research were finally starting to crystallize, literally in the form of cyro-EM structures of the NPC1 protein.
Ironically, one of those strands goes back more than 15 years to the nascent NPC researcher community and the early days of the Parseghian Foundation. This strand faded in the translational era of cyclodextrin — the ten years spanning 2004 to 2014 — but is now regaining favor. Sequence homology comparisons between NPC1 and other proteins performed in the late 1990s showed that NPC1 belongs to an ancient family of membrane pore proteins powered by proton motive forces called RND permeases. Members of the RND permease family are present in bacteria, that’s how ancient we’re talking. It’s true that transmembrane proteins can be more challenging to study than soluble proteins, and it’s true that NPC1 isn’t a classical enzyme, and it’s true that NPC1 is multifunctional. So while all those facts do make NPC1 objectively hard to study in the lab, evolution has been clearly and insistently pointing us in one particular direction. We’ll see next year whether evolution was right all along.
“Heart” image courtesy of the Parseghian Medical Research Fund